You’re out of free articles.
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
Sign In or Create an Account.
By continuing, you agree to the Terms of Service and acknowledge our Privacy Policy
Welcome to Heatmap
Thank you for registering with Heatmap. Climate change is one of the greatest challenges of our lives, a force reshaping our economy, our politics, and our culture. We hope to be your trusted, friendly, and insightful guide to that transformation. Please enjoy your free articles. You can check your profile here .
subscribe to get Unlimited access
Offer for a Heatmap News Unlimited Access subscription; please note that your subscription will renew automatically unless you cancel prior to renewal. Cancellation takes effect at the end of your current billing period. We will let you know in advance of any price changes. Taxes may apply. Offer terms are subject to change.
Subscribe to get unlimited Access
Hey, you are out of free articles but you are only a few clicks away from full access. Subscribe below and take advantage of our introductory offer.
subscribe to get Unlimited access
Offer for a Heatmap News Unlimited Access subscription; please note that your subscription will renew automatically unless you cancel prior to renewal. Cancellation takes effect at the end of your current billing period. We will let you know in advance of any price changes. Taxes may apply. Offer terms are subject to change.
Create Your Account
Please Enter Your Password
Forgot your password?
Please enter the email address you use for your account so we can send you a link to reset your password:
AI has already changed weather forecasting forever.
It’s been a wild few years in the typically tedious world of weather predictions. For decades, forecasts have been improving at a slow and steady pace — the standard metric is that every decade of development leads to a one-day improvement in lead time. So today, our four-day forecasts are about as accurate as a one-day forecast was 30 years ago. Whoop-de-do.
Now thanks to advances in (you guessed it) artificial intelligence, things are moving much more rapidly. AI-based weather models from tech giants such as Google DeepMind, Huawei, and Nvidia are now consistently beating the standard physics-based models for the first time. And it’s not just the big names getting into the game — earlier this year, the 27-person team at Palo Alto-based startup Windborne one-upped DeepMind to become the world’s most accurate weather forecaster.
“What we’ve seen for some metrics is just the deployment of an AI-based emulator can gain us a day in lead time relative to traditional models,” Daryl Kleist, who works on weather model development at the National Oceanic and Atmospheric Administration, told me. That is, today’s two-day forecast could be as accurate as last year’s one-day forecast.
All weather models start by taking in data about current weather conditions. But from there, how they make predictions varies wildly. Traditional weather models like the ones NOAA and the European Centre for Medium-Range Weather Forecasts use rely on complex atmospheric equations based on the laws of physics to predict future weather patterns. AI models, on the other hand, are trained on decades of prior weather data, using the past to predict what will come next.
Kleist told me he certainly saw AI-based weather forecasting coming, but the speed at which it’s arriving and the degree to which these models are improving has been head-spinning. “There's papers coming out in preprints almost on a bi-weekly basis. And the amount of skill they've been able to gain by fine tuning these things and taking it a step further has been shocking, frankly,” he told me.
So what changed? As the world has seen with the advent of large language models like ChatGPT, AI architecture has gotten much more powerful, period. The weather models themselves are also in a cycle of continuous improvement — as more open source weather data becomes available, models can be retrained. Plus, the cost of computing power has come way down, making it possible for a small company like Windborne to train its industry-leading model.
Founded by a team of Stanford students and graduates in 2019, Windborne used off-the-shelf Nvidia gaming GPUs to train its AI model, called WeatherMesh — something the company’s CEO and co-founder, John Dean, told me wouldn’t have been possible five years ago. The company also operates its own fleet of advanced weather balloons, which gather data from traditionally difficult-to-access areas.
Standard weather balloons without onboard navigation typically ascend too high, overinflate, and pop within a matter of hours (thus becoming environmental waste, sad!). Since it’s expensive to do launches at sea or in areas without much infrastructure, there’s vast expanses of the globe where most balloons aren’t gathering any data at all.
Satellites can help, of course. But because they’re so far away, they can’t provide the same degree of fidelity. With modern electronics, though, Windborne found it could create a balloon that autonomously changes altitude and navigates to its intended target by venting gas to descend and dropping ballast to ascend.
“We basically took a lot of the innovations that lead to smartphones, global satellite communications, all of the last 20 years of progress in consumer electronics and other things and applied that to balloons,” Dean told me. In the past, the electronics needed to control Windborne’s system would have been too heavy — the balloon wouldn’t have gotten off the ground. But with today’s tiny tech, they can stay aloft for up to 40 days. Eventually, the company aims to recover and reuse at least 80% of its balloons.
The longer airtime allows Windborne to do more with less. While globally there are more than 1,000 conventional weather balloons launched every day, Dean told me, “We collect roughly on the order of 10% or 20% of the data that NOAA collects every day with only 100 launches per month.” In fact, NOAA is a customer of the startup — Windborne already makes millions in revenue selling its weather balloon data to various government agencies.
Now, with a potentially historic hurricane season ramping up, Windborne has the potential to provide the most accurate data on when and where a storm will touch down.
Earlier this year, the company used WeatherMesh to run a case study on Hurricane Ian, the Category 5 storm that hit Florida in September 2022, leading to over 150 fatalities and $112 billion in damages. Using only weather data that was publicly available at the time, the company looked at how accurately its model (had it existed back then) would have tracked the hurricane.
Very accurately, it turns out. Windborne’s predictions aligned neatly with the storm’s actual path, while the National Weather Service’s model was off by hundreds of kilometers. That impressed Khosla Ventures, which led the company’s $15 million Series A funding round earlier this month. “We haven’t seen meaningful innovation in weather since The Weather Channel in the 90s. Yet it’s a $100 billion market that touches essentially every industry,” Sven Strohband, a partner and managing director at Khosla Ventures, told me via email.
With this new funding, Windborne is scaling up its fleet of balloons as it prepares to commercialize. The money will also help Windborne advance its forecasting model, though Dean told me robust data collection is ultimately what will set the company apart. “In any kind of AI industry, whoever has the top benchmark at any given time, it’s going to fluctuate,” Dean said. “What matters is the model plus the unique datasets.”
Unlike Windborne, the tech giants with AI-based weather models — including, most recently, Microsoft — aren’t gathering their own data, instead drawing solely on publicly accessible information from legacy weather agencies.
But these agencies are starting to get into the game, too. The European Centre for Medium-Range Weather Forecasts has already created its own AI-based model, the Artificial Intelligence/Integrated Forecasting System, which it runs in parallel to its traditional model. NOAA, while a bit behind, is also looking to follow suit.
“In the end, we know we can't rely on these big tech companies to just keep developing stuff in good faith to give to us for free,” Kleist told me. Right now, many of the top AI-based weather models are open source. But who knows if that will last? “It's our mission to save lives and property. And we have to figure out how to do some of this development and operationalize it from our side, ourselves,” Kleist said, explaining that NOAA is currently prototyping some of its own AI-based models.
All of these agencies are in the early stages of AI modeling, which is why you likely haven’t noticed weather predictions making a pronounced leap in accuracy as of late. It’s all still considered quite experimental. “Physical models, the pro is we know the underlying assumptions we make. We understand them. We have decades of history of developing them and using them in operational settings,” Kleist told me. AI-based models are much more of a black box, and there’s questions surrounding how well they will perform when it comes to predicting rare weather events, for which there might be little to no historical data for the model to reference.
That hesitation might not last long, though. “To me it’s fairly obvious that most of the forecasts that would actually be used by users in the future will come from machine learning models,” Peter Dueben, head of Earth systems modeling at the European Centre for Medium Range Weather Forecasting, told me. “If you just want to get the weather forecast for the temperature in California tomorrow, then the machine learning model is typically the better choice,” he added.
That increased accuracy is going to matter a lot, not just for the average weather watcher, but also for specific industries and interest groups for whom precise predictions are paramount. “We can tailor the actual models to particular sectors, whether it's agriculture, energy, transportation,” Kleist told me, “and come up with information that's going to be at a very granular, specific level to a particular interest.” Think grid operators or renewable power generators who need to forecast demand or farmers trying to figure out the best time to irrigate their fields or harvest crops.
A major (and perhaps surprising) reason this type of customization is so easy is because once AI-based weather models are trained, they’re actually orders of magnitude cheaper and less computationally intensive to run than traditional models. All of this means, Kleist told me, that AI-based weather models are “going to be fundamentally foundational for what we do in the future, and will open up avenues to things we couldn't have imagined using our current physical-based modeling.”
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
On betrayed regulatory promises, copper ‘anxiety,’ and Mercedes’ stalled EV plans
Current conditions: Typhoon Wipha is barrelling through southern China, making its way across the mainland after pummeling Hong Kong with heavy rain • More than 60 million Americans are facing heat alerts as temperatures surge • The unusually warm 21-degree Fahrenheit temperature recorded at Summit Station in Greenland is just a few degrees off a record high.
EPA Administrator Lee ZeldinKevin Lamarque-Pool/Getty Images
The Environmental Protection Agency announced plans on Friday afternoon to shut down its research arm and fire hundreds of biologists, chemists, toxicologists, and other scientists whose work helps determine safe pollution levels for regulations. The announcement comes after months of denials from EPA administrator Lee Zeldin that he planned to close the division in question, the Office of Research and Development, which studies the threat from climate change, toxic chemicals, and air and water pollution on human health, and funds university research programs.
The closure comes as part of deep job cuts at the agency. In a statement on Friday, Zeldin said the more than 500 layoffs — which, combined with voluntary buyouts, will slash the EPA’s workforce by nearly one-quarter compared to January’s numbers — would save taxpayers nearly $750 million. The nation’s biggest chemical manufacturers’ lobby agreed, arguing to NPR that the cuts would “ensure American taxpayer dollars are being used efficiently and effectively.” But environmentalists warned that the cuts would “not only cripple EPA’s ability to do its own research, but also to apply the research of other scientists.”
Shares in non-Chinese producers of graphite surged on Friday after the Trump administration slapped new anti-dumping duties of 93.5% on imports of the key mineral for batteries, the Financial Times reported. Combined with existing tariffs on Chinese materials, the new trade levies total more than 160%, according to the consultancy Benchmark Minerals. In response, the stock price for Australia-listed Syrah Resources, the world’s largest non-Chinese graphite miner and the developer behind a key Inflation Reduction Act-funded project in Louisiana, shot up 22%. Canada’s Nouveau Monde Graphite spiked 26%. The dual-listed Australian-U.S. producer Novonix surged 15%.
Not all of President Donald Trump’s mineral tariffs are causing excitement for U.S. allies. Earlier this month, the White House announced 50% tariffs on copper to begin in August, but it has yet to clarify whether those tariffs will apply to refined metal, semi-refined products, or copper ore. The uncertainty is causing “anxiety,” Máximo Pacheco, the chairman of Chile’s state-owned copper miner, told the FT. As Heatmap’s Matthew Zeitlin wrote when the tariffs were first announced, they have the potential to “provide renewed impetus to expand copper mining in the United States. But tariffs can happen in a matter of months. A copper mine takes years to open — and that’s if investors decide to put the money toward the project in the first place.”
Regulators in Virginia last week ordered electricity and natural gas provider Dominion to lay out a clearer blueprint for meeting the state’s legally-enshrined carbon-free electricity targets. But the State Corporation Commission still accepted the monopoly utility’s plans to build out more fossil fuel generation, Canary Media reported.
The Virginia Clean Economy Act, passed in 2020, requires Dominion to generate 100% of its electricity from carbon-free sources by 2045. The accepted plan runs up to 2039, leaving just six years to sort out the details of decarbonization. The regulators cautioned that “acceptance does not express approval.” While the statement stopped short of calling into question a proposed 944-megawatt gas complex just south of Richmond, Virginia’s capital, the commission said it would debate plans for another roughly 5 gigawatts of gas-burning power plants before approving construction..
British energy giant BP is selling off its U.S. onshore wind business as the Trump administration appears ready to smother the industry. On Friday, New York-based developer LS Power said it agreed to buy BP’s share of 10 wind projects totaling 1,700 megawatts of capacity. As part of the deal, LS Power plans to fold the wind projects into its renewable energy subsidiary, Clearlight Energy, increasing its fleet to 4,300 megawatts.
BP’s exit comes as the Trump administration has vowed to crack down on the expansion of wind and solar power in the U.S. Trump has long personally opposed wind energy, dating back to his unsuccessful fight against turbines erected near his golf course in Scotland before entering politics. Last week, Heatmap’s Jael Holzman reported on a memo from the Department of the Interior calling for political reviews of essentially all solar and wind developments in the U.S. This would at minimum stretch out the already challenging development timeline for projects, a problem especially as developers rush to qualify for federal tax credits.
Mercedes-Benz is pumping the brakes on U.S. production of its EQ line of electric vehicles as the Trump administration winds down federal tax credits to support purchases of battery-powered cars. The German automaker told InsideEVs that, by the start of September, it planned to temporarily pause assembly lines of all variants of its EQE and EQS sedans and SUVs that are either located in the U.S. or producing vehicles bound for the American market. The manufacturer is no longer taking orders from dealers for the cars.
Reviewers had criticized the EQ models for lacking the quality and sophistication of similar gas-powered lines of Mercedes vehicles. Even before Republicans in Congress rolled back the federal government’s landmark $7,500 tax credit for EVs, Mercedes faced trouble finding buyers. Sales of the EQS sedan and EQS SUV were down 52% in the U.S. in 2024 compared to the previous year. China’s biggest electric automakers, meanwhile, are racing to build factories in Brazil, the largest car market in South America.
Like tiny winged Magellans measuring barely an inch in size, the Bogong moth of Australia regularly travels more than 600 miles using celestial navigation, according to a new study in Nature. “The moths really are using a view of the night sky to guide their movements,” a researcher told Euronews.
From the Inflation Reduction Act to the Trump mega-law, here are 20 years of changes in one easy-to-read cheat sheet.
The landmark Republican reconciliation bill, which President Trump signed on July 4, has shattered the tax credits that served as the centerpiece of the country’s clean energy and climate policy.
Starting as soon as October, the law — which Trump has dubbed the One Big Beautiful Bill Act — will cut off incentives for Americans to install solar panels, purchase electric vehicles, or make energy efficiency improvements to their homes. It’s projected to raise household energy costs while increasing America’s carbon emissions by 190 million metric tons a year by 2030, according to the REPEAT Project at Princeton University.
The loss of these incentives will in part offset the continuation of tax cuts that largely benefit wealthy Americans. But the law as a whole won’t come close to paying for those cuts in their entirety. The legislation is expected to swell federal deficits by nearly $3.8 trillion over the next 10 years, according to the Tax Foundation, a nonpartisan think tank. This explosive deficit expansion could make it more difficult for the Federal Reserve to cut interest rates, possibly further constraining energy development.
President Trump has described the law as ending Democrats’ “green new scam,” and conservative lawmakers have celebrated the termination of Biden-era energy programs. The law is particularly devastating for programs encouraging electric vehicle sales, as well as wind and solar energy deployment.
But the act is more complicated than a simple repeal of Democrats’ 2022 Inflation Reduction Act. In one case, Trump’s big law ends a federal energy incentive that has been in place, in some form, since the 1990s. In others, Republicans have tied up existing energy incentives with new restrictions, regulations, and red tape.
Some parts of the IRA have even remained intact. GOP lawmakers opted to preserve Biden’s big expansion of incentives to support nuclear energy and advanced geothermal development. That said, the Trump administration could still gut these tax credits by making them effectively unusable through executive action.
It can be confusing to keep the One Big Beautiful Bill Act’s many changes to federal energy law in your head — even for experts. That’s why Heatmap News is excited to publish this new reference “cheat sheet”on the past, present, and future of federal energy tax credits, compiled by an all-star collection of analysts and researchers.
The summary takes each clean energy-related provision in the U.S. tax code and summarizes how (and whether) it existed in the 2000s and 2010s, how the Inflation Reduction Act changed it, and how the new OBBBA will change it again. It was compiled by Shane Londagin, a policy advisor at the think tank Third Way; Luke Bassett, a former Biden administration official and Senate Energy committee staffer; Avi Zevin, a former Biden official and a partner at the energy law firm Roselle LLP; and researchers at the REPEAT Project, an energy analysis group at Princeton University. (Note that I co-host the podcast Shift Key with Jesse Jenkins, who leads the REPEAT Project.)
You can find the full summary below.
On presidential proclamations, Pentagon pollution, and cancelled transmission
Current conditions: Over 1,000 people have evacuated the region of Seosan in South Korea following its heaviest rainfall since 1904 • Forecasts now point toward the “surprising return” of La Niña this fall • More than 30 million people from Louisiana through the Appalachians are at risk of flash flooding this weekend due to an incoming tropical rainstorm.
The Hugh L. Spurlock Generating Station in Maysville, Kentucky.Jeff Swensen/Getty Images
President Trump on Thursday signed four proclamations allowing certain highly polluting industries to bypass regulations established by the Biden administration. In addition to chemical manufacturers that help produce semiconductors and medical device sterilizers, the proclamations singled out coal-fired power plants and taconite iron ore processing facilities for two years of exemptions. Taconite is a low-grade iron ore primarily mined in the Upper Peninsula of Michigan and northern Minnesota, which is then processed for use in the production of iron and steel. Trump justified the move by arguing that compliance with the current emissions rule for coal-fired power plants raises the “unacceptable risk” of shutdowns, “eliminating thousands of jobs, placing our electrical grid at risk, and threatening broader, harmful economic and energy security effects,” while the iron processing emissions rule “risks forcing shutdowns, reducing domestic production, and undermining the nation’s ability to supply steel for defense, energy, and critical manufacturing.”
The proclamations allow industries to comply with the Environmental Protection Agency standards that predate former President Joe Biden’s tenure. Trump justified the pause by claiming the former administration had mandated compliance with “standards that rely on emissions-control technologies that have not been demonstrated to work.” Researchers have previously found that air pollutants related to coal power plants cause nearly 3,000 attributable deaths per year. Taconite iron ore processing facilities produce harmful acid gases, including hydrogen chloride and hydrogen fluoride, as well as mercury, which have been linked to numerous adverse health effects.
Separately, the House passed Trump’s $9 billion rescissions package late last night, which includes cuts to international climate, energy, and environmental programs like the Clean Technology Fund. Republicans Brian Fitzpatrick of Pennsylvania and Mike Turner of Ohio joined Democrats in objecting to the bill. Trump is expected to sign the package Friday. An additional rescissions package is expected “soon.”
The Pentagon’s 2026 budget will enable the Department of Defense’s planet-warming emissions to grow by an additional 26 megatons, or about the equivalent of 68 gas power plants, a new analysis by the Climate and Community Institute found. The U.S. military was already the single largest institutional polluter in the world due to its “vast global operations — from jet fuel consumption and overseas deployments to domestic base maintenance,” as well as its manufacturing of weapons and vehicles, the think tank notes. With the passage of the One Big Beautiful Bill Act, the Pentagon’s budget will exceed $1 trillion in 2026, representing a 17% increase over 2024. Its emissions, in turn, could grow to the point that if the DOD were its own country, it’d be the 38th largest polluter in the world, producing more CO2 emissions than the Netherlands, Bangladesh, or Venezuela. But “the Pentagon’s true climate impact will almost certainly be worse” than what the researchers found, The Guardian notes, “as the calculation does not include emissions generated from future supplemental funding such as the billions of dollars appropriated separately for military equipment for Israel and Ukraine in recent years.”
Get Heatmap AM directly in your inbox every morning:
New York’s Public Service Commission decided Thursday against moving forward with a major transmission project that would have had the capacity to deliver at least 4,770 megawatts of offshore wind power to New York City by the early 2030s. The commissioners said they were unable to justify “charging ratepayers for the multibillion-dollar project when feds are stymying” offshore wind, New York Focus’ Colin Kinniburgh reported on Bluesky. “We will continue to press forward regarding infrastructure needs for offshore wind in the future once the federal government resumes leasing and permitting for wind energy generation projects,” PSC chair Rory Christian said.
The canceled Public Policy Transmission Need determination was not specific to a particular offshore wind project, but rather was intended to match New York’s general offshore wind ambitions when it was approved in 2023. But as Heatmap has previously reported, Trump’s crusade against offshore wind has been a “worst case scenario” for the industry since day one, and, per ABC News 10, effectively “eliminates any reason for building new power lines in the first place.”
Microsoft has inked a deal to purchase 4.9 million metric tons of durable carbon dioxide removal from Vaulted Deep, a waste management startup, for an undisclosed amount. The companies boasted that the deal, which runs through 2038, represents “the second-largest carbon removal deal to date.” Vaulted Deep, an Xprize Carbon runner-up, diverts organic waste from landfills and incinerators by injecting it into wells thousands of feet underground using fracking technologies, which it says ensures over 1,000 years of durability, TechCrunch reports. Since Vaulted’s launch in the summer of 2023, the Houston-based company has removed 18,000 metric tons of carbon dioxide. Microsoft, meanwhile, has slipped behind its 2020 goal to remove more carbon from the atmosphere than it generates by the end of the decade due to its rush to build out data centers.
The Environmental Protection Agency’s reorganization and downsizing are set to continue, with the agency offering another round of buyouts and early retirements to staffers in offices it aims to restructure, Politico reports. Among the affected offices are the Office of Enforcement and Compliance Assurance, which the EPA said it seeks to tweak to “better address pollution problems that impact American communities by re-aligning enforcement with the law to deliver economic prosperity and ensure compliance with agency regulations,” as well as the Office of Land and Emergency Management, which works on Superfund and disaster response issues. The Office of Research and Development, the Office of Mission Support, and the Office of the Chief Financial Officer are also affected.
Separately, in a preliminary decision earlier this week, the agency moved to block the state of Colorado from closing its six remaining coal-fired power plants by 2031. Colorado was attempting to codify the retirement dates in its Regional Haze Plan, which is typically used to protect the air quality of federal wilderness and national parks; however, the EPA rejected the proposal, according to CPR News. “We believe that the Clean Air Act does not give anybody the authority to shut down coal generation plants against the owner’s will,” Cyrus Western, the administrator of EPA Region 8, said. Jeremy Nichols, a senior advocate for the Center of Biological Diversity’s environmental health program, claimed the EPA’s move shows the limits of what climate-conscious states can do on their own. “We may have state rules, but they won't be federally approved,” Nichols told CPR.
“There are so many developers and so many projects in so many places of the world that there are examples where either something goes wrong with a project or a developer doesn’t follow best practices. I think those have a lot more staying power in the public perception of renewable energy than the many successful projects that go without a hiccup and don’t bother people.” —Heatmap Pro’s Charlie Clynes, in conversation with Jael Holzman about his new project tracking all of the nation’s county-level restrictions on renewable energy.